Modeling of the Galvanic Anode Cathodic Protection System with
Dynamic Polarization Characteristics
ADNAN MUJEZINOVIĆ, IRFAN TURKOVIĆ, ALIJA MUHAREMOVIĆ
Faculty of Electrical Engineering
University of Sarajevo
Zmaja od Bosne bb, Sarajevo
BOSNIA AND HERZEGOVINA
[email protected], [email protected], [email protected]
SANJA MARTINEZ
Faculty of Chemical Engineering and Technology
University of Zagreb
Trg Marka Marulića 19, Zagreb
CROATIA
SLOBODAN MILOJKOVIĆ
Faculty of Electrical Engineering
University of East Sarajevo
BOSNIA AND HERZEGOVINA
Abstract: - Cathodic protection systems are used for protection of underground or underwater metallic
infrastructure against corrosion. Protection by cahtodic protection system is achieved by polarizing protected
object i.e. by shifting equilibrium potential to more negative value. Value of the electric potential and current
density on the surface of protected object are essential data for evaluation of efficiency of the cathodic
protection system. Value of these parameters can be determined by using numerical techniques. This paper
deals with the mathematical modeling of cathodic protection system when taking into account dynamic
nonlinear polarization characteristics on the electrode surface. Firstly, mathematical model is described in
detail. Numerical procedure presented in this paper is divided in the two parts. First part is the calculation of
distribution of electric potential and current density in spatial domain using direct boundary element method,
and calculation of the time changes of these parameters by finite difference time domain method. Finally,
presented mathematical model was used for calculation of parameters of one geometrically simple cathodic
protection system example.
Key-Words: - Cathodic Protection System, Galvanic Anode, Dynamic Nonlinear Polarization Chracteristics,
Boundary Element Method, Finite Difference Time Domain Method, Newton-Raphson Technique.
1 Introduction Cathodic protection is most widely used technique
for protection of the underground and underwater
metallic structures from corrosion [1]. This
technique is based on the shifting the equilibrium
potential of protected structure to more negative
value. This can be done by connecting the protected
object with additional electrode(s) placed in same
electrolyte (ground or water), which equilibrium
potential is more negative than equilibrium potential
of the protected object [2]. After installation of the
cathodic protection system, electric potential value
on the entire surface of protected structure must be
lower than minimum protection potential value
defined by standards [3]. Also, current density
distribution on the surface of protected structure
should be uniform as possible [4]. Therefore, for
evaluation of efficiency of the cathodic protection
system, electric potential and protection current
density distribution on the protected metallic
structure surface need to be known [5]. These
parameters can be determined by solving Laplace
Adnan Mujezinovic et al.International Journal of Chemistry and Chemical Engineering Systems
http://www.iaras.org/iaras/journals/ijcces
ISSN: 2367-9042 95 Volume 1, 2016
partial differential equation along with the adequate
boundary conditions [1]. This type of the problems
can be easily solved by using numerical techniques.
For modeling of the cathodic protection systems,
boundary element method is mostly used [6-8]. The
main advantage of this method, in comparison to the
other numerical methods, is that it requires
discretization only of boundaries of considered
domain and there is no need for discretization of
infinite boundaries [9]. Additionally, iterative
techniques for solution of nonlinear equations (such
as Newton – Raphson technique) must be included
because nonlinear boundary conditions (nonlinear
polarization characteristics) are given on electrode
surfaces of cathodic protection system.
In some situations, these nonlinear boundary
conditions can change over the time (dynamic
nonlinear polarization characteristics) [10-12].
These time changes of boundary conditions affect
the electric potential and current density distribution
in the system. Therefore, for modeling of the
cathodic protection system with dynamic nonlinear
polarization characteristics, some modifications in
mathematical model must be done. In this paper,
this problem is solved by using combined boundary
element method and finite difference time domain
method.
2 Mathematical Model As previously mentioned, the purpose of modeling
of the cathodic protection system is to determine the
distribution of protection current density and electric
potential on the surface of the protected structure in
order to verify the effectiveness of the system.
Electric potential distribution of cathodic protection
system can be determined by solving Laplace partial
differential equation for static current field:
02 (1)
where φ is electric potential and is Nabla
operator.
To obtain a unique solution it is necessary to add
the appropriate boundary conditions on domain
boundaries to a Laplace partial differential equation.
Boundaries that are taken into account when
modeling cathodic protection system are
anode/electrolyte interface and cathode/electrolyte
interface. Boundary conditions applied to these
boundaries are called polarization characteristics
and they are influenced by electrochemical reaction
that takes place on electrode surfaces. Polarization
characteristics represent the correlation between
current density and electric potential on electrode
surfaces and they are nonlinear. When polarization
characteristics change over the time then they are
dynamic polarization characteristics. These changes
of polarization characteristics over the time can be
caused by calcareous and magnesium deposition on
the electrode surface. This phenomenon is
characteristic for the cathode surface [11,13].
Dynamic polarization characteristic on the
cathodic surface is defined by the following
mathematical relation:
FeFe
HHOOC
fj
fjfjj
Fe
Fe
H
H
*
2
2
*2
222
10
10*
(2)
where j*O2
is threshold current density of oxygen
reduction, jH2 is current density of the hydrogen
separation, jFe is current density of the metal
dissolution, φ*H2
and φ*Fe are equilibrium potentials
for corresponding electrochemical reactions, βH2 and
βFe are Tafel’s coefficients, φ is potential difference
of interface metal/electrolyte and fO2, fH2
and fFe are
time-dependent factors. For all three partial
electrochemical reactions, the time-dependent
factors have a following form:
RR
t
R
t
R efef
1, (3)
where index R represents a corresponding
electrochemical reaction.
On the anode surface, polarizing characteristic is
defined by the following mathematical relation:
1102
a
a
Oa jj
(4)
where jO2 is current density corresponding to the
oxygen reduction, φa is equilibrium potential of
galvanic anode and βa is Tafel slope of galvanic
anode.
Graphical representation of the polarization
characteristics for anode and cathode surface is
given on the Figure1. Polarization characteristics are
Adnan Mujezinovic et al.International Journal of Chemistry and Chemical Engineering Systems
http://www.iaras.org/iaras/journals/ijcces
ISSN: 2367-9042 96 Volume 1, 2016
constructed using relations (2) and (4) and data from
literature [13].
Figure 1. Polarization characteristics of the
electrode surfaces
For calculation of the electric potential and current
density distribution of the cathodic protection
system with dynamic polarization characteristic,
numerical procedure based on boundary element
method and finite difference time domain method
was used. This numerical procedure can be divided
into two parts. First part is calculation of electric
potential and current density distribution in spatial
domain at t = 0 using boundary element method and
second part for calculation of electric potential and
current density change over the time using finite
difference time domain method.
2.1 BEM formulation in spatial domain Boundary element method for calculation of the
electric potential and current density distribution in
spatial domain is based on solving the integral field
equation. Integral field equation can be obtained by
applying Green's symmetric identity on Laplace
partial differential equation (1). Then integral field
equation have the following form:
dpjqpG
dqpGnpqqc
,
,
(5)
where p is position of the source point, q is position
of the field point, Γ is boundary of the domain, φ(p)
is electric potential of source point p, φ(q) is electric
potential of field point q, c(q) is a constant, n
is
normal unit vector, ρ is resistivity of surrounding
electrolyte, j(p) is current density at source point p,
φ∞ is constant potential at infinite boundary and
G(p,q) is Green’s function whose form for 2D plan
parallel problems is as follows:
qp
qpG
1
ln2
1,
(6)
For problems with infinite boundaries, Gauss
boundary condition must be added [14]:
0
dpj (7)
In this way it is ensured that there is no flow of
current on infinite boundary, and by using Gauss
boundary condition there is no need for
discretization of infinite boundary.
In the process of calculation of electric potential
and current density distribution using boundary
element method, it is necessary to discretize domain
boundary on appropriate number of boundary
elements. After discretization of domain boundaries,
field integral equation (5) is solved over every
element. After application of the Collocation
method at the point and numerical integration,
solution of the electric potential and current density
can be obtained for each node of all boundary
elements by solving following matrix equation [15]:
jGH (8)
where {φ} is vector of unknown electric potentials
in colocation points, {j} is vector of current
densities in colocation points, [H] and [G] are
matrix of influence coefficients of system geometry.
Since nonlinear boundary conditions are
specified, to solve the matrix equation (8) it is
necessary to apply an iterative technique. In this
paper, matrix equation (8) is solved by using
Newton – Raphson iterative technique proposed in
[11,12].The first step in solving a matrix equation
(8) by Newton - Raphson technique is expansion of
the vector current density in the Taylor series, as
follows:
kk
kk
jjj
1
1 (9)
where k is number of the current iteration, [∂j/∂φ] is
Jacobian matrix, and {∆φ} is vector of electric
potential increments between the two adjacent
iteration, and can be written as:
1 kkk (10)
By including matrix equations (9) and (10) in the
equation (8), following equation can be written
[11,12]:
Adnan Mujezinovic et al.International Journal of Chemistry and Chemical Engineering Systems
http://www.iaras.org/iaras/journals/ijcces
ISSN: 2367-9042 97 Volume 1, 2016
11
kkk HjG
jGH
(11)
From previous matrix equation vector of electric
potential increments in k – th iteration can be
calculated. Further, this vector is used in matrix
equations (9) and (10) for calculation of current
densities and electric potentials in same iteration.
Iterative cycle is repeated until a convergence is
achieved.
Since, this calculation can be time consuming
some modifications are done in this paper in order
to accelerate calculation procedure. In matrix
equations (9) and (11) Jacobian matrix is used. This
matrix is rectangular matrix whose all elements
outside the main diagonal equal to zero. Therefore,
instead of memorizing a square matrix, non – zero
elements of Jacobian matrix are stored in form of
vector. Also, GMRES iterative technique was used
for calculation of matrix equation (11) in each
iteration of Newton – Raphson method.
2.2 FTDT formulation On cathode surface dynamic (time – varying)
polarization characteristics are assumed. Therefore,
to calculate change of the electric potential and
current density over the time previously presented
mathematical model needs to be expanded.
nn
n
n
nn tt
jjjj
11
1
(12)
where n is time step at which current density is
calculating, [∂j/∂t] is matrix of first derivatives of
current density function in the time, {∆t} is vector
of time steps and {∆φ}n is vector of electric
potentials increments in n-th time step and can be
written as:
1 nnn (13)
From equation (4) it is noticeable that
polarization characteristic of anode is time
independent, therefore in matrix equation (12)
elements of matrix [∂j/∂t] that corresponds to
colocation points placed on anode surface are equal
to zero.
In matrix equation (12) two vectors are
unknown. By including matrix equations (12) and
(13) in the matrix equation (8), matrix equation with
one unknown vector can be obtained, as follows:
1
1
1
1
nn
n
nn
n
Htt
jG
jGj
GH
(14)
By solving previous matrix equation vector of
electric potential increments in new time step is
obtained. Further, this vector is used in matrix
equations (12) and (13) for calculation of current
densities and electric potentials in same time step.
3 Case Study Application of the previously presented
mathematical model is demonstrated on example
given on the Figure 2. Radius of anode is 1 m while
radius of cathode is 5 m. Both, anode and cathode
are placed in the electrolyte with resistivity 25 Ωcm.
Electrolyte
Cathode
rc
Anodera
d 0°
90°
180°
270°
270°
90°
0°
180°
Figure 2. Geometry of analyzed galvanic anode
catohdic protection system
Below are presented results of the calculation of
electric potential and current density on the cathode
and anode surface of example given in Figure 2.
Distribution of electric potential on the cathode
surface is given on the Figure 3.
Figure 3. Electric potential distribution on the
cathode surface
Results given on the Figure 3 indicate that areas
that are closer to the anode (the nearest point of the
cathode to the anode is located at 180 °) have a
Adnan Mujezinovic et al.International Journal of Chemistry and Chemical Engineering Systems
http://www.iaras.org/iaras/journals/ijcces
ISSN: 2367-9042 98 Volume 1, 2016
lower value of electric potential. This further means
that this area is more protected than the other parts
of the cathode. Also, it can be noticed that there is a
change of the electric potential to negative side in
all points of the cathode, which supports the
cathodic protection of surfaces against corrosion.
Change of the current density distribution on the
cathode surface in the discrete time steps is given on
the Figure 4.
Figure 4. Current density distribution on the cathode
surface over the time
From results given on the Figure 4 it is
noticeable that current density is higher (in absolute
sense) at areas that are closer to the anode. Also,
increase of the current density over the time (also in
absolute sense) is higher in areas that are closer to
the anode.
Electric potential and current density distribution
on the anode surface over the time is given on the
Figures 5 and 6, respectively.
Figure 5. Electric potential distribution on the anode
surface over the time
Figure 6. Current density distribution on the anode
surface over the time
On anode surface it can be noticed that the
electrical potential is more positive on area where
the anode surface is closer to the cathode (the
nearest point of the anode to the cathode is located
at 0 °). Also, on this area current density is the
highest.
Change of the current intensity in analyzed
cathodic protection system, caused by dynamic
polarization characteristics are given on the Figure
7.
Figure 7. Change of the current intensity in system
over the time
From results given on the Figure 7 it can be
noted that current intensity decreases over the time.
Also, highest change of the current intensity is at
first day, while after two days current intensity is
constant.
4 Conclusion Design of modern cathodic protection system
requires application of very accurate and precise
methods for calculating of relevant parameters. The
reason for this is the fact that the boundary
conditions on the electrode surface of the system are
nonlinear and in some situations, such as deposition
of calcareous and magnesium on the cathode
surface, and may be time-varying (dynamic). In this
Adnan Mujezinovic et al.International Journal of Chemistry and Chemical Engineering Systems
http://www.iaras.org/iaras/journals/ijcces
ISSN: 2367-9042 99 Volume 1, 2016
paper, mathematical model for calculation of the
electric potential and current density distribution in
galvanic anode cathodic protection system with
dynamic nonlinear polarization characteristics is
presented. Presented mathematical model is based
on combination of boundary element method and
finite difference time domain method. Application
of presented mathematical model was demonstrated
on one 2D simple geometry problem.
Acknowledgment This paper was supported by the Ministry of
Education and Science of the Federation of Bosnia
and Herzegovina.
References:
[1] L. Lazzari, P.Pedeferri, Cathodic Protection, 1st
edition, Polipress Milano, 2006.
[2] B. Jarić, A. Rešetić, Corrosion:
Electrochemical fundamentals and cathodic
protection, Korexpert, Zagreb, 2003.
[3] W. J. Santos, J. A. F. Santiago, J. C. F. Telles,
Using the Gaussian function to simulate
constant potential anodes in multiobjective
optimization of cathodic protection systems,
Engineering Analysis with Boundary Elements,
Vol. 73, 2016, pp. 35-41.
[4] A. Muharemović, H. Zildžo, E. Letić,
Modelling of protective potential distribution in
a cathodic protection system using a cupled
BEM/FEM method, 30TH International
conference on Boundary Elements Method and
Other Reduction Methods, BEM/MRM 30,
Maribor, Slovenia, 2008.
[5] I. Turković, Contribution to the study of the
influence of nonlinear voltage and current
distribution to optimization of the protection
length in cathodic protection domain, PhD
Dissertation, Sarajevo 2010.
[6] C. Liu, A. Shankar, M. E. Orazem, P. D.
Riemer, Numerical simulations for cathodic
protection of pipelines, in Underground
pipeline corrosion - Detection, analysis and
prevention, Edited by M. E. Orazem,
Woodhead Publishing, 2014.
[7] J. F. Fan, S. N. R. Pakalapati, T. V. Nguyen, R.
E. White, R. B. Griffin, Mathematical
Modelling of Cathodic Protection Using the
Boundary Element Method with a Nonlinear
Polarization Curve, Journal of The
Electrochemical Society, Vol. 139, No. 7, 1992,
pp. 1932-1936.
[8] A. Mujezinović, A. Muharemović, I. Turković,
A. Muharemović, Calculation of the Protective
Current Density Distribution of a Cathodic
Protection System with Galvanic Anodes in
Terms of Double-layer Electrolyte, 34TH
International conference on Boundary
Elements Method and Other Reduction
Methods, BEM/MRM, Croatia, Split, 2012.
[9] S. Martinez, Evaluation of the Uniform Current
Density Assumption in Cathodic Protection
System with Close Anode to Cathode
Arrangement, Materials and Corrosion, Vol.
61, No. 4, 2010, pp. 338 – 342.
[10] K. Nisancioglu, Predicting the Time
Dependence of Polarization on Cathodically
Protected Steel in Seawater, Corrosion, Vol.
43, No. 2, 1987, pp. 100 - 111.
[11] J. A. F. Santiago, J. C. F. Telles, On Boundary
Elements for Simulation of Cathodic Protection
System with Dynamic Polarization Curves,
International Journal for Numerical Methods
in Engineering, Vol. 40, No. 14, 1997, pp.
2611 – 2627.
[12] J. A. F. Santiago, J. C. F. Telles, A Solution
Technique for Cathodic Protection System with
Dynamic Boundary Conditions by the
Boundary Element Method, Advances in
engineering software, Vol. 340, No. 9, 1999,
pp. 663 – 671.
[13] S. L. Carson, M. E. Orazem, Time – dependent
polarization behaviour of pipeline grade steel in
low ionic strength environments, Journal of
Applied Electrochemistry, Vol. 29, No. 6, 1999,
pp. 707 – 721. [14] C. A. Brebbia, J. C. F. Telles, L. C. Wrobel,
Boundary Element Techniques Theory and
Applications in Engineering, Berlin /
Heidelberg, Springer-Verlag, 1984.
[15] A. Mujezinović, I. Turković, S. Martinez, S.
Milojković, Modelling of the Cathodic
Protection System with Dynamic Non-Linear
Polarization Characteristics, XXV International
Conference on Information, Communication
and Automation Technologies (ICAT), Sarajevo
2015.
Adnan Mujezinovic et al.International Journal of Chemistry and Chemical Engineering Systems
http://www.iaras.org/iaras/journals/ijcces
ISSN: 2367-9042 100 Volume 1, 2016